Re: [PATCH] nvme: add cond_resched() to nvme_complete_batch()

[Niklas Cassel]

On Wed, May 03, 2023 at 05:42:22PM -0600, Keith Busch wrote:
> On Wed, May 03, 2023 at 10:28:53AM -0600, Keith Busch wrote:
> > On Tue, May 02, 2023 at 08:54:12PM +0800, jiweisun126@xxxxxxx wrote:
> > > According to the above two logs, we can know the nvme_irq() cost too much
> > > time, in the above case, about 4.8 second. And we can also know that the
> > > main bottlenecks is in the competition for the spin lock pool->lock.
> > 
> > I have a fix that makes pool->lock critical section much shorter and constant
> > time. It was staged in mm tree for months, but mysteriously it's missing...
> 
> Well this is certainly odd: the commit I'm referring to is in Linus' log, but
> the file it modified is not updated. What kind of 'git' madness is this??
> 
> https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=a4de12a032fa6d0670aa0bb43a2bf9f812680d0f
> 
> https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/log/mm/dmapool.c
> 

Hmm, this is very interesting indeed.

It looks like:
a4de12a032fa ("dmapool: link blocks across pages") matches
https://patchwork.kernel.org/project/linux-mm/patch/20230126215125.4069751-12-kbusch@xxxxxxxx/
so everything good.

However, the following commit:
2d55c16c0c54 ("dmapool: create/destroy cleanup")
does not match what was on the list:
https://patchwork.kernel.org/project/linux-mm/patch/20230126215125.4069751-13-kbusch@xxxxxxxx/



It kind of looks like
2d55c16c0c54 ("dmapool: create/destroy cleanup") is actually
a revert of a4de12a032fa ("dmapool: link blocks across pages"),
plus some additional changes. Scary.



Additionally, it seems like the fix:
https://patchwork.kernel.org/project/linux-mm/patch/20230221165400.1595247-1-kbusch@xxxxxxxx/
was never merged.


Something is fishy here...

(Perhaps the maintainer did a git revert instead of applying the fix...
and accidentally squashed the revert with a proper commit...
But even that does not make sense, as there simply seems to be too many
lines changed in 2d55c16c0c54 ("dmapool: create/destroy cleanup") for that
to be the case...)


Kind regards,
Niklas